Data captured during evolutions performed by aerial vehicles prior to one or more missions, and data regarding outcomes of the missions, may be used to train a machine learning system to predict data regarding an outcome of a mission of an aerial vehicle based on the performance of the aerial vehicle during one or more evolutions. The data may be captured by sensors provided aboard an aerial vehicle, or in association with a testing facility, and may include data captured during both pre-flight and/or in-flight evolutions performed by the aerial vehicle. The evolutions may include any pre-flight operation of motors, propellers and/or control surfaces, or any other components, as well as the in-flight operation of such components. If a machine learning system determines that a mission is unlikely to succeed, the mission may be canceled, delayed until further inspections may be performed, or assigned to another aerial vehicle.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A system comprising: a landing area; a first plurality of sensors associated with the landing area, wherein the first plurality of sensors comprises at least one of an acoustic sensor, a current sensor, an imaging device, a load sensor, a vibration sensor, a voltage sensor, a tachometer or a thermometer, and wherein the landing area is within an operating range of each of the first plurality of sensors; and a computer server connected to at least one network, wherein the computer server is in communication with each of the first plurality of sensors, and wherein the computer server is configured to at least: cause a first aerial vehicle to perform at least a first evolution while at least a portion of the first aerial vehicle is grounded on the landing area, wherein the first evolution comprises an operation of at least one component of the first aerial vehicle while at least the portion of the first aerial vehicle is grounded on the landing area; determine, by at least some of the first plurality of sensors, first data regarding the first aerial vehicle during at least the first evolution; provide at least some of the first data to a machine learning system as a first input, wherein the machine learning system is trained to predict an outcome of a first mission of the first aerial vehicle based at least in part on data captured by at least one of the first plurality of sensors; receive a first output from the machine learning system based at least in part on the first input; determine, based at least in part on the first output, that the first aerial vehicle is likely to complete a first mission; and cause the first aerial vehicle to take off from the landing area and perform the first mission.
2. The system of claim 1 , wherein the computer server is further configured to at least: receive, from the first aerial vehicle over the at least one network, second data regarding the first aerial vehicle during at least the first evolution, wherein the second data is captured by at least one of a second plurality of sensors provided aboard the first aerial vehicle; and provide at least some of the second data to the machine learning system as a second input, wherein the machine learning system is trained to predict the outcome of the first mission based at least in part on data captured by at least one of the first plurality of sensors and data captured during at least the first evolution, and wherein the first output is received from the machine learning system based at least in part on the first input and the second input.
3. The system of claim 1 , wherein the computer server is further configured to at least: during the first mission, receive, from the first aerial vehicle over the at least one network, second data regarding the first aerial vehicle during at least a portion of the first mission, wherein the second data is captured by at least one of a second plurality of sensors provided aboard the first aerial vehicle; provide at least some of the second data to the machine learning system as a second input, wherein the machine learning system is trained to predict the outcome of the first mission based at least in part on data captured during at least one of the first evolution or the second evolution; receive a second output from the machine learning system based at least in part on the second input; determine, based at least in part on the second output, that the first aerial vehicle is unlikely to complete the first mission; and cause the first aerial vehicle to abort the first mission.
4. A method comprising: causing a first aerial vehicle to execute a first operation of at least one component of the first aerial vehicle, wherein the first operation is executed with the first aerial vehicle grounded or airborne not greater than a predetermined altitude within an operating range of at least a first sensor associated with a ground-based facility; capturing first data during the first operation by at least the first sensor; providing at least some of the first data to a machine learning system as a first input, wherein the machine learning system is trained to predict data regarding an outcome of a mission to be performed by an aerial vehicle based at least in part on data captured during at least one operation of the aerial vehicle; receiving, from the machine learning system, a first output based at least in part on at least the first input; and predicting second data regarding a first outcome of a first mission of the first aerial vehicle based at least in part on the first output.
5. The method of claim 4 , wherein the first operation comprises at least one of: operating a propulsion motor of the first aerial vehicle at a first predetermined rotating speed; changing a speed of the propulsion motor from the first predetermined rotating speed to a second predetermined rotating speed; operating a control surface of the first aerial vehicle from a first position to a second position; extending or retracting a landing component of the first aerial vehicle; retrieving or releasing an object by a payload engagement system of the first aerial vehicle; operating an imaging device provided on the first aerial vehicle; or supplying power to at least one of the propulsion motor, the control surface, the landing component, the payload engagement system or the imaging device.
6. The method of claim 4 , wherein the first aerial vehicle is tethered to a ground surface during the first operation.
7. The method of claim 4 , further comprising: capturing third data during the first operation by at least a second sensor associated with at least one of the first aerial vehicle, a second aerial vehicle or the ground-based facility, wherein providing at least some of the first data to the machine learning system as the first input comprises: providing at least some of the first data and at least some of the third data to the machine learning system as the first input.
8. The method of claim 7 , wherein the second sensor is at least one tachometer, wherein the third data comprises at least an operating speed of the propulsion motor of the first aerial vehicle.
9. The method of claim 4 , wherein the first sensor is at least one acoustic sensor, and wherein the first data comprises at least one of a frequency or a sound pressure level of a sound emitted by at least a portion of the first aerial vehicle during the first operation.
10. The method of claim 7 , wherein the second sensor is at least one of a current sensor or a voltage sensor, and wherein the third data comprises at least one of a current or a voltage supplied to the at least one component during the first operation.
11. The method of claim 4 , wherein the first sensor is at least one temperature sensor, and wherein the first data comprises at least an operating temperature of at least a portion of the at least one component during the first operation.
12. The method of claim 4 , wherein the first sensor is at least one imaging device, and wherein the first data comprises at least one image of the at least one component captured by the at least one imaging device during the first operation.
13. The method of claim 4 , wherein predicting the second data regarding the first outcome of the first mission of the first aerial vehicle comprises: determining whether the first mission is likely to succeed based at least in part on the second data; and one of: in response to determining that the first mission is likely to succeed, causing the first aerial vehicle to embark on the first mission; or in response to determining that the first mission is not likely to succeed, at least one of: causing the first aerial vehicle to land at one of an origin of the first mission or an alternate location; canceling the first mission; or causing a second aerial vehicle to embark on the first mission.
14. The method of claim 4 , further comprising: prior to causing the first aerial vehicle to execute the first operation of the at least one component of the first aerial vehicle, causing at least a second aerial vehicle to execute a second operation of at least one component of at least the second aerial vehicle; capturing third data during the second operation by at least a second sensor; determining fourth data regarding at least a second outcome of at least a second mission of the second aerial vehicle; and training the machine learning system to associate at least the third data with at least the fourth data.
15. The method of claim 4 , wherein the machine learning system is one of: a nearest neighbor method; an artificial neural network; a conditional random field; a factorization method; a K-means clustering analysis; a log likelihood similarity measure; a cosine similarity measure; a latent Dirichlet allocation; or a latent semantic analysis.
16. The method of claim 4 , further comprising: determining at least one attribute of the first mission, wherein predicting the second data regarding the first outcome of the first mission of the first aerial vehicle comprises: predicting the second data regarding the first outcome of the first mission of the first aerial vehicle based at least in part on the first output and the at least one attribute of the first mission.
17. A system comprising: at least a first sensor associated with one of a ground-based testing facility or an airborne testing facility; at least one data store; and at least one computer processor in communication with at least the first sensor and connected to at least one network, wherein the at least one computer processor is configured to at least: capture, by the first sensor, data during performance of a plurality of testing evolutions by each of a plurality of aerial vehicles; receive, from each of the plurality of aerial vehicles, data regarding outcomes of missions performed by the plurality of aerial vehicles following the performance of the plurality of testing evolutions by the plurality of aerial vehicles; train a machine learning system to associate at least some of the data captured during the performance of the plurality of testing evolutions with at least some of the data regarding the outcomes of the missions; capture, by the first sensor, first data during performance of at least a first testing evolution by a first aerial vehicle; provide at least some of the first data to the trained machine learning system as a first input; receive a first output from the trained machine learning system; and predict an outcome for a first mission of the first aerial vehicle based at least in part on the first output, wherein the first sensor is not provided aboard the first aerial vehicle or any of the plurality of aerial vehicles.
18. The system of claim 17 , wherein the plurality of testing evolutions comprises a sequence of operations of each of a plurality of powered elements aboard an aerial vehicle, wherein the plurality of powered elements includes at least one propulsion motor, at least one control surface, at least one landing component or at least one aspect of a payload engagement system, wherein the first data is captured during performance of the plurality of testing evolutions by the first aerial vehicle, and wherein the first testing evolution is one of the operations of one of the plurality of powered elements aboard the aerial vehicle.
19. The system of claim 17 , wherein the first aerial vehicle is tethered to at least one surface at the ground-based testing facility or within an operating range of the first sensor from the airborne testing facility during the performance of the first testing evolution.
20. The system of claim 17 , wherein the at least one computer processor is further configured to at least: capture, by a second sensor provided aboard the first aerial vehicle, second data during the performance of at least the first testing evolution by the first aerial vehicle; and provide at least some of the second data to the trained machine learning system, wherein the first input comprises the at least some of the first data and the at least some of the second data.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
October 20, 2017
October 20, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.